Methodology: How to compare Presidencies over time

This is a wonky methodology post that some readers may want to skip. I am including it because I need to use it in another post today, and I anticipate I may need it again in the future.

Economists often cringe when people try to compare economic outcomes across presidencies. Economists generally think the right way to compare short-term macroeconomic performance is to compare business cycles, and the timing of business cycles never matches up perfectly with the electoral calendar.

Budget analysts have an additional problem: the federal fiscal year begins October 1, so in a Presidential transition year one budget year spans two Presidencies.

There’s a third unsolvable problem, which is that many economic policies work with a long lag. President Bush’s first year macroeconomy was in part defined by the after effects of the tech bubble collapse of 2000, before he took office. Similarly, President Obama’s first year was heavily shaped by a financial crash that happened four months before he took office.

Still, the policy debate and political process requires that we find a fair and objective way to compare economic and fiscal outcomes across Presidencies.

For budget data I think the best way is to look at the nine budget years spanned by a President’s two terms, and to measure that ninth year mid-year, when he leaves office. (For a one-term President we would examine five budget years.)

This is easiest with a concrete example. CY stands for Calendar Year and is measured January 1 – December 31. FY stands for (federal) Fiscal Year and is measured October 1 – September 30.

  • In CY 2000, his last year in office, President Clinton negotiated with Congress on the spending and tax bills for FY 2001, which began October 1, 2000, almost four months before President Bush took office.
  • In January 2001, CBO estimated the budget outlook for FY 2001 based on the laws signed into effect for FY 2001 by President Clinton.
  • President Bush worked with Congress to quickly enact the 2001 tax cuts. That law significantly changed the FY 2001 budget outlook. So even though President Clinton determined much of the FY 2001 budget, we need to hold President Bush responsible for FY 2001 as well since (a) he served for eight of its 12 months and (b) he had a big effect on it. This is true even though President Clinton shaped most of the FY 2001 budget.
  • President Bush was in office for all of FY 2002 through FY 2008. That’s seven more years. We’re up to eight.
  • FY 2009 began October 1, 2008. TARP was enacted in the first few days of that fiscal year, and we committed more than $300 B in the last four months of the Bush presidency, which also were the first four months of FY 2009.
  • President Bush therefore should be held accountable for FY 2009 as it was projected when he left office in January, 2009. Lucky for us, CBO releases its annual Economic and Budget Outlook with a new baseline each January, so we have an independent estimate of that ninth year before any Obama policies take effect.

Therefore the best and fairest way to measure fiscal policy over the Bush Administration is to look at nine budget years:

  • FY 2001 actuals, even though President Clinton was responsible for much of its shape;
  • and FY 2002 – FY 2008 actuals;
  • and FY 2009 as estimated by CBO in their January 2009 Economic and Budget Outlook.

The same holds true for any other President. While President Bush should be held accountable for the projected FY 2009 budget outlook (his “ninth” year) when he left office, President Obama should be held responsible for the final FY 2009 outcome, including any policies he signed into law and any economic changes that occurred in the eight months of that year when he was President.

You can see this works nicely for the Bush-Obama transition:

  • Bush’s TARP spending ($300+ B) goes on Bush’s account, as does the decline in revenues that occurred in FY 2009, during part of which he served as President.
  • Obama’s TARP spending and Obama’s stimulus goes on Obama’s account in the FY 2009 actuals, as does any further weakening of the economy and revenues after Obama took office.

The downside is that policies occurring in the first four months of the first fiscal year of a Presidency are largely shaped by one’s predecessor. That’s undesirable, but to some extent each President “inherits” the sum total of everything that comes before them, and I don’t know any way to avoid this without causing more measurement inequities.

This methodology leads to the slightly surprising result of calculating fiscal policy averages over nine years for a two-term President, and over five years for a one-term President. It also means that boundary years get attributed to two Presidents (albeit measured at different times). This feels a little odd but after some thought it seems to make sense.

This methodology is imperfect but it’s the best I can construct. It’s not fair that Bush’s first measured year was largely shaped before he took office, nor that Obama’s measured first year was largely shaped before he took office. But any other way of measuring Presidencies (like attributing all of FY 2009 to Bush) runs into bigger problems – surely Bush shouldn’t be assigned the budget effects of the Obama stimulus, any more than Clinton should be assigned the budget effects of the Bush tax cut.

If anyone can suggest a way to improve this methodology (and not just point out its flaws, please) I would love to hear it. I want/need something which is objective, rule-based, and fair.

3 responses

  1. Pingback: A decade of spiraling deficits | KeithHennessey.com

  2. Pingback: A decade of spiraling deficits | The Daily Caller - Breaking News, Opinion, Research, and Entertainment

  3. If you take 9 years, how can you avoid double-counting? It appears you must then assign half of the first and last years of a presidency both to that president and to his predecessor and successor. Why not simply count from, e.g., July of the first term until the July following the last? That is, if you don't want to use a weighted average for the first & last terms…

Follow

Get every new post delivered to your Inbox.

Join 6,423 other followers